Postgraduate research opportunities EngD in Machine Learning for reliable automated inspection of aerospace composites (with an enhanced stipend)

The increasing use of advanced engineering materials such as Carbon Fibre Reinforced Plastic (CFRP) composites in the aerospace industry offers enormous ecological and financial benefits as the reduced final weight of aerostructures directly helps save on fuel consumption. This trend can be seen in modern civil aircrafts such as the Airbus A380 where 25% of the weight is consisting of composites that are largely manufactured and supplied by Spirit AeroSystems, as the technology and market leader for years. CFRP exhibits superior mechanical tensile properties in preferred directions where the application loadings are expected to be the highest owing to the composition of carbon fibres and resin. However, a number of manufacturing defects such as pores, delamination, lack of bonding, in-plane/through-thickness fibre waviness, and changes in the fibre volume fraction can occur after molding. Since the manufactured CFRP components are safety-critical and should be of the highest integrity to be used in airframes, non-destructive testing (NDT) based on methods such as Phased Array Ultrasound Testing (PAUT) is an essential post-manufacturing stage for certification.

PAUT probes and controllers allow for individual transmit/receive of the probe array elements enabling electronic beamforming, focusing, and steering within the target material. This introduces improved inspection coverage and reliability as compared to conventional UT systems. However, manual inspection of typical large-sized aircraft components made of CFRPs such as wing covers, pressure bulkheads, fuselage, and flaps is quite a slow process creating a bottleneck in the entire production cycle. The recent advances in the deployment of industrial robots for NDT, and particularly for UT of CFRPs [1], however, have alleviated some of the hurdles for the inspection speed. Although high-speed inspection systems that sometimes reach 500 mm/s acquiring 10,000 frames/s have obvious benefits, the enormous data obtained through these should be managed and processed by intelligent algorithms to truly reach the full potential of the automated inspection.

The encoded PAUT data generated by these scans are in form amplitude scans (A-scans) which corresponds to the amplitude versus time response of transmit-receive by each element/sub-aperture. Different projections such as B-scan, C-scan, and D-scan of the volumetric data can be produced to efficiently detect and characterize the potential bulk defects. This project will explore the implementation of automated PAUT data interpretation for CFRPs through two approaches: I) developing a low latency Deep Neural Network (DNN) to analyse the A-scan data on the fly, while the scan is being performed, for geometrical feature recognition, automated gating of time series, and defect detection, and II) developing a Multitask Network (MN) [2] for image analysis, detection of geometrical features/defects on each B-scan, D-scan, and C-scan projection, and cross-validation of findings at the combination stage. The real-time DNN applied to the A-scan data will serve to provide warnings for defects flagged during the inspection while the MN, with a potentially improved learning through different related tasks empowered by the multi-view analysis of the data, should be able to detect the defects with higher confidence.

The project is relevant to the many advanced industrial sectors such as Aerospace, Defence, Automotive & general High-Value Manufacturing striving to bring autonomy to their production/ inspection processes using machine learning.

Aligned with the financial commitment from the industrial partner, the project scope is centred around the current NDE demands of Spirit AeroSystems with the target to a) develop and deliver more industry-focused NDE solutions to promote the partner's and UK's business growth, and b) to introduce development program for the student, where highly demanded skills by the industry, access to a network of NDE experts in academia and industry, access to the state-of-the-art research facilities, and specialized NDE training can be offered to the student.

The project will make extensive use of the £2.5 million cutting-edge Sensor Enabled Automation & Control Hub (SEARCH) hosting several advanced industrial robots and NDE equipment at the Centre for Ultrasonic Engineering (CUE) at the University of Strathclyde. The student will have access to and will work closely with the Aerospace Innovation Centre (AIC) established by Spirit AeroSystems at their Prestwick manufacturing facility and NMIS facilities in Renfrew.

The student will work within an internationally renowned and growing team of diverse and multi-disciplinary researchers and engineers, physicists, and mathematicians and will receive a full NDE training package through FIND CDT and a university training for working with advanced industrial KUKA robots, different NDE controllers and sensor technologies.

[1] Mineo, Carmelo, et al. "Flexible integration of robotics, ultrasonics and metrology for the inspection of aerospace components." AIP conference proceedings. Vol. 1806. No. 1. AIP Publishing LLC, 2017.

[2] Zhou, Yue, et al. "Multi-task learning for segmentation and classification of tumors in 3D automated breast ultrasound images." Medical Image Analysis 70 (2021): 101918.